The Agent Skills Directory

[SAFE]: The skill defines a workflow for validating app metadata, event drafts, and review responses. The instructions and the referenced MCP tool validate_ai_companion_data are consistent with the primary purpose of the skill and the vendor's provided services.
[SAFE]: No malicious patterns such as remote code execution, credential exfiltration, or obfuscation were detected in the instructions or metadata.
[PROMPT_INJECTION]: The skill is designed to process untrusted external data, including app metadata and review drafts, which constitutes a surface for indirect prompt injection. * Ingestion points: Untrusted data enters the agent context through the data parameter of the validation tool and the content of the active page. * Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the processed data. * Capability inventory: The skill utilizes the validate_ai_companion_data tool; no file system access, subprocess execution, or direct network operations are present. * Sanitization: The skill does not describe any specific validation or sanitization of the input data prior to processing.

metadata-validation